Victor Riparbelli was a Copenhagen computer science graduate obsessed with science fiction and machine learning when he came across academic research showing how a person's face could be puppeteered in real time. In 2017, he co-founded Synthesia with Steffen Tjerrild and two academics – UCL's Professor Lourdes Agapito and TUM's Professor Matthias Niessner, both of whom have since scaled back to advisory and research roles – on the bet that most corporate video didn't need a camera, a crew or a studio at all.

Early experiments in film and multilingual dubbing generated headlines, but the money arrived when Synthesia aimed its AI avatars at enterprise training, onboarding and internal comms, like training videos. The platform now counts 70 per cent of the FTSE 100 among its customers.

In January 2026, Synthesia raised a $200m Series E, led by Alphabet's GV, with Nvidia's NVentures, Accel, NEA, Hedosophia and Air Street Capital participating – by then it had reached $150m in annual recurring revenue. Growth has come with scrutiny. Synthesia's avatars have been reportedly misused in state propaganda videos. In answer, Synthesia closed the offending accounts and modified its content moderation guidelines.

The company’s long-term proposition now is that enterprise buyers will pick the safest synthetic video vendor over the most permissive one. Meanwhile, it’s also figuring out how its pioneering tech can also be used in a world of hybrid work, where humans work alongside AI agents.